Aim: This study aims to explain variations in adolescent smoking status using predictor variables measured during childhood. Unlike most of the extant research to identify predictors of adolescent smoking, this study will use a prospective cohort design that spans the years from childhood to adolescence, and thus will span the pre-initiation to nicotine dependent phases of cigarette smoking. Moreover, several of the predictor variables selected for study (i.e., parenting style, anti-smoking socialization by parents, and children's susceptibility to smoking) have never before been measured during childhood and then used to predict smoking status in adolescence. Method: Previous research by the principal investigator provides the foundation for this application. From 1994 through 1996 the investigator completed a study of initiation of smoking in a cohort of 632 children who were in grades 3 through 5 during the study years. The proposed study will utilize telephone survey research methods to obtain 8-year post-baseline data from this established cohort. Variables: The primary independent variables are: (1) general parenting practices; (2) anti-smoking socialization by parents; (3) children's competence development; (4) children's susceptibility to smoking; (5) children's initial experience with smoking. The outcome variables are specified as either established smoking, current smoking, or phase of smoking. Hypotheses: Among the main effect hypotheses to be tested are: (i) Compared with peers exposed to other styles of parenting, children exposed to authoritative parenting will have significantly lower odds of established smoking in adolescence. (ii) There will be a significant inverse association between amount of exposure to anti-smoking socialization during childhood and the odds of established smoking during adolescence. (iii) The higher the susceptibility to smoking during childhood the greater the odds of established smoking in adolescence. Additional hypotheses posit main effects, mediator effects, or moderator effects of the independent variables on smoking outcomes in adolescence. The results of this study will be of value because they will help determine whether childhood is indeed a critical period for preventing smoking during adolescence, and they will specify which factors merit attention in future smoking prevention programs for children.